analiza aiqbench
This research exposes a critical vulnerability in voice authentication systems by demonstrating how adversarial attacks can manipulate imperceptible audio frequencies to bypass deep learning-based anti-spoofing defenses. The study reveals that current static detection models fail against adaptive threats, achieving near-perfect success rates against standalone speaker verification and combined security frameworks. These findings underscore the urgent need for dynamic, context-aware defense mechanisms capable of evolving alongside sophisticated AI-generated audio threats.
Spectral Masking and Interpolation Attack (SMIA): A Black-box Adversarial Attack against Voice Authentication and Anti-Spoofing Systems
arXiv:2509.07677v5 Announce Type: replace-cross Abstract: Voice Authentication Systems (VAS) use unique vocal characteristics for verification. They are increasingly integrated into high-security sectors such as banking and healthcare. Despite their improvements using deep learning, they face severe vulnerabilities from sophisticated threats like deepfakes and adversarial attacks. The emergence of realistic voice cloning complicates detection, as systems struggle to distinguish authentic from synthetic audio. While anti-spoofing countermeasures (CMs) exist to mitigate these risks, many rely on static detection models that can be bypassed by novel adversarial methods, leaving a critical security gap. To demonstrate this vulnerability, we propose the Spectral Masking and Interpolation Attack (SMIA), a novel method that strategically manipulates inaudible frequency regions of AI-generated audio. By altering the voice in imperceptible zones to the human ear, SMIA creates adversarial samples that sound authentic while deceiving CMs. We conducted a comprehensive evaluation of our attack against state-of-the-art (SOTA) models across multiple tasks, under simulated real-world conditions. SMIA achieved a strong attack success rate (ASR) of at least 82% against combined VAS/CM systems, at least 97.5% against standalone speaker verification systems, and 100% against countermeasures. These findings conclusively demonstrate that current security postures are insufficient against adaptive adversarial attacks. This work highlights the urgent need for a paradigm shift toward next-generation defenses that employ dynamic, context-aware frameworks capable of evolving with the threat landscape.
Abstract:Voice Authentication Systems (VAS) use unique vocal characteristics for verification. They are increasingly integrated into high-security sectors such as banking and healthcare. Despite their improvements using deep learning, they face severe vulnerabilities from sophisticated threats like deepfakes and adversarial attacks. The emergence of realistic voice cloning complicates detection, as systems struggle to distinguish authentic from synthetic audio. While anti-spoofing countermeasures (CMs) exist to mitigate these risks, many rely on static detection models that can be bypassed by novel adversarial methods, leaving a critical security gap. To demonstrate this vulnerability, we propose the Spectral Masking and Interpolation Attack (SMIA), a novel method that strategically manipulates inaudible frequency regions of AI-generated audio. By altering the voice in imperceptible zones to the human ear, SMIA creates adversarial samples that sound authentic while deceiving CMs. We conducted a comprehensive evaluation of our attack against state-of-the-art (SOTA) models across multiple tasks, under simulated real-world conditions. SMIA achieved a strong attack success rate (ASR) of at least 82% against combined VAS/CM systems, at least 97.5% against standalone speaker verification systems, and 100% against countermeasures. These findings conclusively demonstrate that current security postures are insufficient against adaptive adversarial attacks. This work highlights the urgent need for a paradigm shift toward next-generation defenses that employ dynamic, context-aware frameworks capable of evolving with the threat landscape.
Sound (cs.SD); Artificial Intelligence (cs.AI)
arXiv:2509.07677 [cs.SD]
(or arXiv:2509.07677v5 [cs.SD] for this version)
https://doi.org/10.48550/arXiv.2509.07677 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 311 cuvinte
Subiecte conexe
Adv-TGD: Adversarial Text-Guided Diffusion for Face Recognition Impersonation Attacks
arXiv · 12 iun. 2026
Transferable Spatial Temporal Coherence Adversarial Attack on Black-Box Vision Language Models for Autonomous Driving
arXiv · 7 oct. 2026
All you need to break LLMs are Black-Box, Adapting, Efficient, Transferable, Harmful, Applicable ... Attacks
arXiv · 7 oct. 2026
Co-Evolving Zero-Day Jamming: Adaptive Attack Synthesis and Graph Attention-Based Online Detection
arXiv · 22 sept. 2026